Before we begin explaining our progress, we are going to do a small introduction of what our project is all about. For the project, we have decided to create a robotic arm where we would like to design it so that it has the same movements as humans have. We have looked at many different robotic arm projects, but none of them have free degree of movement. This means that their movement was limited to the X and Y axis, whereas our arm will have both the X and Y axis, but also the Z axis which would rotate the arm, just like humans do.
A nice extension for this project would be to use a computer vision technology like the Microsoft Kinect sensor to track the users arm movement and reproduce the same behaviour in the robot. This can then have plenty of applications, like especially in the industry. There are already some kind of solutions existing to control a robotic arm using a real one but they are usually based on a direct interaction with some kinds of commands. Usually the user uses a device like a joystick. The main point of using the Kinect would be to get more freedom movement and especially it would be more intuitive to use.
Since the project is not limited only to Arduino, we have decided to use something different to experiment new systems. The software we use is called Urbiscript which is developped by Gostai and is a special software used for programming robotics. The servos that we are going to use are Dynamixel AX-12 (http://www.crustcrawler.com/motors/AX12/index.php). These servos have a library for Urbiscript called bioloid which we are going to use so that we can control the servos in the way we want. Although the project might seem simple, just because we only use 4 servos, it is actually more challenging than it looks, since it needs to be built in the right way in order to mimic the human movements.
The servos are pretty good because they have a very good torque and also they have built in sensors where they would switch off if something wrong happens which is one of the initial things we wanted to have for our arm because if we move our arm and gets stuck into something, it will switch off the servos with the most pressure on them so they do not break. The servos also have a built in potentiometer that allows us to find the position of the servos. However we will explain more as we go along on the weekly progress of the arm. Last thing we would like to mention is that since we have many different assignment to complete, we are not going to spend more than 2 or 3 days per week for the project and we would also like to mention that each member has contributed equally by designing different parts of the project.
For the first week, we haven't really done much, but what we did was quite important for the whole project because we did research on different techniques of robotic arm creation, and we did the planning for our own arm. Below are a few links of Youtube videos where people have created their own robotic arms. Usually, robotic arms are built from a bottom base so they can easily pick up objects from the ground. But our project is made to mimic the behaviours of a human arm, which is attached on a top position. That is a main difference between our arm and the ones show in the videos.
http://www.youtube.com/watch?v=1c4t9ou3ReA&feature=related <<<<< here is another robotic arm, which seems to be built from scratch and uses a wii remote control to be moved around, but it is very similar to the above link and doesn't seem very efficient and it still does not provide the same degree of movement as a humans arm.
Since we were not satisfied with these arms we have decided to create our own paper based robotic arm to test the arm movements. It helped us to figure out how to create the link between the motors and to know how many of them we would need. We needed to create a ball join used to reproduce the shoulder's motions and a simple rotation for the elbow. Creating a paper based version helped a lot especially to realise the ball join. A ball join is based on 3 degrees rotation freedom usually named X,Y and Z. To create it we needed 3 motors.
Below are a few images of what the paper based arm looks like.
So far, that is what we did during the first week on this project. We have done some research on robotic arms, but none did what we really wanted. The good solutions were usually created by researchers with the help of fundings so they were usually quite expensive. For the others, they were based on a ground attached arm so they did not completely fit with our idea.
Week 1 Progress
Introduction
Before we begin explaining our progress, we are going to do a small introduction of what our project is all about. For the project, we have decided to create a robotic arm where we would like to design it so that it has the same movements as humans have. We have looked at many different robotic arm projects, but none of them have free degree of movement. This means that their movement was limited to the X and Y axis, whereas our arm will have both the X and Y axis, but also the Z axis which would rotate the arm, just like humans do.
A nice extension for this project would be to use a computer vision technology like the Microsoft Kinect sensor to track the users arm movement and reproduce the same behaviour in the robot. This can then have plenty of applications, like especially in the industry. There are already some kind of solutions existing to control a robotic arm using a real one but they are usually based on a direct interaction with some kinds of commands. Usually the user uses a device like a joystick. The main point of using the Kinect would be to get more freedom movement and especially it would be more intuitive to use.
Since the project is not limited only to Arduino, we have decided to use something different to experiment new systems. The software we use is called Urbiscript which is developped by Gostai and is a special software used for programming robotics. The servos that we are going to use are Dynamixel AX-12 (http://www.crustcrawler.com/motors/AX12/index.php). These servos have a library for Urbiscript called bioloid which we are going to use so that we can control the servos in the way we want. Although the project might seem simple, just because we only use 4 servos, it is actually more challenging than it looks, since it needs to be built in the right way in order to mimic the human movements.
The servos are pretty good because they have a very good torque and also they have built in sensors where they would switch off if something wrong happens which is one of the initial things we wanted to have for our arm because if we move our arm and gets stuck into something, it will switch off the servos with the most pressure on them so they do not break. The servos also have a built in potentiometer that allows us to find the position of the servos. However we will explain more as we go along on the weekly progress of the arm. Last thing we would like to mention is that since we have many different assignment to complete, we are not going to spend more than 2 or 3 days per week for the project and we would also like to mention that each member has contributed equally by designing different parts of the project.
For the first week, we haven't really done much, but what we did was quite important for the whole project because we did research on different techniques of robotic arm creation, and we did the planning for our own arm. Below are a few links of Youtube videos where people have created their own robotic arms. Usually, robotic arms are built from a bottom base so they can easily pick up objects from the ground. But our project is made to mimic the behaviours of a human arm, which is attached on a top position. That is a main difference between our arm and the ones show in the videos.
http://www.youtube.com/watch?v=1qG9H0mAezw <<<<< the robotic arm which is on that link can be bought for around £35 from
http://www.play.com/Gadgets/Gadgets/4-/10713210/Robot-Arm/Product.html?_%24ja=tsid:11518%7Ccat:10713210%7Cprd:10713210, however that wouldn't satisfy us since it is already built and it doesn't really do what we want it to do!
http://www.youtube.com/watch?v=1c4t9ou3ReA&feature=related <<<<< here is another robotic arm, which seems to be built from scratch and uses a wii remote control to be moved around, but it is very similar to the above link and doesn't seem very efficient and it still does not provide the same degree of movement as a humans arm.
Since we were not satisfied with these arms we have decided to create our own paper based robotic arm to test the arm movements. It helped us to figure out how to create the link between the motors and to know how many of them we would need. We needed to create a ball join used to reproduce the shoulder's motions and a simple rotation for the elbow. Creating a paper based version helped a lot especially to realise the ball join. A ball join is based on 3 degrees rotation freedom usually named X,Y and Z. To create it we needed 3 motors.
Below are a few images of what the paper based arm looks like.
So far, that is what we did during the first week on this project. We have done some research on robotic arms, but none did what we really wanted. The good solutions were usually created by researchers with the help of fundings so they were usually quite expensive. For the others, they were based on a ground attached arm so they did not completely fit with our idea.